rageshmookoth / startbot Goto Github PK

Create a chatbot with Flask, Python and Rasa NLU

using supervised training

Backend:
step1) Startup csv data file from Kaggle
step2) Use Jupyter notebook for data analysis
step3) Create training data json file using rasa nlu trainer: https://rasahq.github.io/rasa-nlu-trainer/
step4) mkdir startbot
step5) mkdir startbot/data to store the  startup data json file
       data files : data/startup-data.json
step6) Cretae a config file that uses spacy for NLP, model directory to store the trained model
       config.json
step7) Train the chatbot using rasa_nlu and config by running the  following command
       python -m rasa_nlu.train -c config.json
       This will create a model in the directory specified (models) in the config files with the timestamp
step8) Set up http server to run with the config file and latest model
       python -m rasa_nlu.server -c config.json 
step9) To setup the knowledge base for the startbot create a knowledge file with responses using python dictionary.
        knowledge.py

Frontend:
Flask application
Step1) Create templates folder and the following templates: layout.html, index.html
Step2) Create static folder and css and js sub folders: css/chat-interface.css, js/bind.js
Step3) Use bootstrap, googlefontapi
Step4) Create app.py : main flask application
       python app.py to begin